chemical composition and droplet size distribution of ... · 1environment research institute,...

Atmos. Chem. Phys., 17, 9885–9896, 2017https://doi.org/10.5194/acp-17-9885-2017© Author(s) 2017. This work is distributed underthe Creative Commons Attribution 3.0 License.

Chemical composition and droplet size distribution of cloud at thesummit of Mount Tai, ChinaJiarong Li1, Xinfeng Wang1, Jianmin Chen1,2,3, Chao Zhu1, Weijun Li1, Chengbao Li2, Lu Liu1, Caihong Xu1,Liang Wen1, Likun Xue1, Wenxing Wang1, Aijun Ding3, and Hartmut Herrmann2,4

1Environment Research Institute, School of Environmental Science and Engineering, Shandong University,Ji’nan, 250100, China2Shanghai Key Laboratory of Atmospheric Particle Pollution and Prevention, Department of Environmental Scienceand Engineering, Institute of Atmospheric Sciences, Fudan University, Shanghai, 200433, China3Institute for Climate and Global Change Research, School of Atmospheric Sciences, Nanjing University,Nanjing, 210023, Jiangsu, China4Leibniz Institute for Tropospheric Research (TROPOS), Atmospheric Chemistry Department (ACD),Permoserstr. 15, 04318, Leipzig, Germany

Correspondence to: Jianmin Chen ([email protected], [email protected])and Hartmut Herrmann ([email protected])

Received: 8 January 2017 – Discussion started: 16 February 2017Revised: 11 June 2017 – Accepted: 25 July 2017 – Published: 23 August 2017

Abstract. The chemical composition of 39 cloud samplesand droplet size distributions in 24 cloud events were inves-tigated at the summit of Mt. Tai from July to October 2014.Inorganic ions, organic acids, metals, HCHO, H2O2, sul-fur(IV), organic carbon, and elemental carbon as well as pHand electrical conductivity were analyzed. The acidity of thecloud water significantly decreased from a reported value ofpH 3.86 during 2007–2008 (Guo et al., 2012) to pH 5.87in the present study. The concentrations of nitrate and am-monium were both increased since 2007–2008, but the over-compensation of ammonium led to an increase in the meanpH value. The microphysical properties showed that clouddroplets were smaller than 26.0 µm and most were in therange of 6.0–9.0 µm at Mt. Tai. The maximum droplet num-ber concentration (Nd) was associated with a droplet size of7.0 µm. High liquid water content (LWC) values could facili-tate the formation of larger cloud droplets and broadened thedroplet size distribution. Cloud droplets exhibited a stronginteraction with atmospheric aerosols. Higher PM2.5 levelsresulted in higher concentrations of water-soluble ions andsmaller sizes with increased numbers of cloud droplets. Thelower pH values were likely to occur at higher PM2.5 con-centrations. Clouds were an important sink for soluble ma-terials in the atmosphere. The dilution effect of cloud water

should be considered when estimating concentrations of sol-uble components in the cloud phase.

1 Introduction

Cloud droplets are formed by the condensation of water va-por on anthropogenic and natural aerosols that serve as cloudcondensation nuclei (CCN). Clouds significantly affect theearth’s radiation budget and they are also responsible for thechanges in regional and global climate (Miles et al., 2000).Cloud events can transport pollutants, promote acid depo-sition, change the meteorological conditions, modify localenvironmental features and affect the fate of several atmo-spheric species via chemical and physical processes (Mooreet al., 2004).

The chemical properties of clouds are initially determinedby CCN (Sun et al., 2010). However, they can be alteredby absorbing chemical components of soluble gases and fur-ther multiphase chemical reactions taking place in the cloudphase (Ravishankara, 1997). Non-precipitating clouds playa more crucial role in ion deposition and aggregation thanprecipitating clouds do (Aleksic et al., 2009). The concentra-tions of soluble compounds and dissolved acids have gener-

Published by Copernicus Publications on behalf of the European Geosciences Union.

9886 J. Li et al.: Chemical composition and droplet size distribution of cloud

ally been reported to be much higher in cloud liquid watercompared with precipitation (Błas et al., 2008; Zapletal etal., 2007; Zimmermann et al., 2003). For example, Sun etal., (2010) found that the concentrations of ammonium, sul-fate and nitrate in cloud water were at least 5.17 times higherthan those in rainwater.

Cloud plays a significant role in scavenging aerosols viadrop deposition (directly or by coalescence into precipita-tion) and in creating new particles and trace gases (Herckeset al., 2002). These processes could influence the distributionand the concentration of pollutants both in the cloud phaseand in the aerosol phase, and also influence the microphysi-cal properties of the clouds (Collett Jr. et al., 2002; Lee et al.,2012; Ogawa et al., 2000). For example, for a given super-saturated condition, an increase in the concentration of CCNwill lead to the formation of small droplets (Borys et al.,2000; Gultepe and Milbrandt, 2007). In addition, the clouddroplet size distribution (CDSD) is prominently determinedby the chemical and physical properties of CCN (Portin etal., 2014; Zipori et al., 2015). Numerous studies have exam-ined the chemical compositions of orographic clouds (Kimet al., 2006; Marinoni et al., 2004; Watanabe et al., 2010),many of which have focused on the size-dependent chemi-cal properties of the clouds (Moore et al., 2004; Schell et al.,1997). However, few studies provide detailed descriptions ofthe interactions between aerosols and the chemical and mi-crophysical properties of clouds.

In this study, cloud samples were collected at the summitof Mt. Tai. It is interesting that the acidity of the cloud waterwas significantly lower than that reported during 2007–2008(Guo et al., 2012; Wang et al., 2011). The causes behind thischange were investigated by examining the chemical compo-sitions of cloud samples at Mt. Tai. We then investigated themicrophysical properties of cloud droplets, including CDSD,LWC, effective diameter (ED) and droplet number concen-tration (Nd). Lastly, we explored the interactions betweencloud droplets and aerosols in the atmosphere.

2 Methods

2.1 Site description and sampling

Mt. Tai (117◦13′ E, 36◦18′ N; 1545 m a.s.l.) is a natural andcultural heritage site in China and it is one of the world’sgeoparks. Because the summit of Mt. Tai lacks emissionsof anthropogenic pollution, the pollutants investigated fromthere could accurately represent the characteristics of theregional pollutants in the North China Plain. The high-frequency of local cloud events, especially in summer, makesMt. Tai a favorable site for collecting cloud samples andmonitoring cloud events. Previous research has indicated thatthe clouds at the summit of Mt. Tai are acidic (Wang et al.,2008).

From 24 July to 31 October 2014, a total of 85 cloudsamples associated with 24 cloud events were collected us-ing a single-stage Caltech active strand cloud-water collector(CASCC), as described by Demoz et al. (1996), and 39 cloudsamples were analyzed. The cloud droplets were drawn intothe collector by a fan with a flow rate of 24.5 m3 min−1 andimpacted on six Teflon nets that each contained 102 strandsof 508 µm in diameter. The samples were then guided along agroove at the bottom of the collector and finally collected intoa 500 mL high-density polyethylene cylinder. The theoretical50 % collection efficiency cut-off size of the cloud dropletsis at 3.5 µm. In this study, sampling-time resolution was ad-justed during sampling sessions in order to ensure that eachsample contained an adequate amount of cloud water for theanalysis (at least 150 mL). The volumes of the samples, thestart and end times of the collection sessions and the num-bers of collected samples were accurately recorded for eachcloud event.

It should be noted that the collector was immediatelyshut down during precipitation to eliminate the interrup-tions caused by rain water. Before each sampling session,the collector was rinsed with high-purity deionized water(≥ 18.2 M�), dried naturally and sealed. Blanks were pre-pared using high-purity deionized water, and then they weretreated and analyzed using the same method as collectedsamples.

2.2 In situ and laboratory analysis

The pH, electrical conductivity and the concentrations ofsulfur(IV), formaldehyde and hydrogen peroxide were mea-sured immediately after sampling. Approximately 10 mL ofeach cloud sample was used to measure the pH and electricalconductivity by using a portable pH meter (model 6350M,JENCO) that was regularly calibrated using standard solu-tions at pH= 4 and pH= 7. Approximately 20 mL of eachcloud sample was filtered using a cellulose acetate filter withpore sizes of 0.45 µm to remove any suspended particulatematter and then the concentrations of sulfur(IV), formalde-hyde and hydrogen peroxide were analyzed in situ to avoidany changes in their concentrations. The measurement meth-ods were described in detail by Collett and colleagues (Col-lett Jr. et al., 1998). For each sample, a 10 mL aliquot wasprepared for trace metal analysis by adding 1 % (v/v) nitricacid and then preserved in a brown glass bottle at 4 ◦C. An-other 10 mL aliquot was prepared to analyze organic acids byadding 0.5 % (v/v) chloroform (to prevent the reproductionof microorganisms) and then storing the solution in a glassbottle at 4 ◦C. The residuals were refrigerated at −20 ◦C forfurther analysis.

The concentrations of eight inorganic ions (Cl−, NO−3 ,SO2−

4 , NH+4 , Na+, K+, Ca2+ and Mg2+) in each sam-ple were measured using an ion chromatograph (Dionex,ICS-90) and the concentrations of four organic acids (ac-etate, formate, oxalate and lactate) were measured using ion-

Atmos. Chem. Phys., 17, 9885–9896, 2017 www.atmos-chem-phys.net/17/9885/2017/

J. Li et al.: Chemical composition and droplet size distribution of cloud 9887

Table 1. Summary of the chemical compositions for cloud samples collected at Mt. Tai during July to October 2014.

Species Units No. samples Min Max VWMb Percentage

pH – 39 3.80 6.93 5.87 –Electrical conductivity µS cm−1 39 44.9 813.5 169.0 –Na+ mg L−1 39 BDLa 2.9 0.9 0.56NH+4 mg L−1 39 5.2 143.3 34.2 21.41K+ mg L−1 39 BDLa 6.5 1.3 0.81Mg2+ mg L−1 39 0.2 3.0 0.7 0.44Ca2+ mg L−1 39 BDLa 39.2 5.9 3.69Cl− mg L−1 39 0.6 11.7 2.9 1.82NO−3 mg L−1 39 2.7 538.5 56.4 35.31SO2−

4 mg L−1 39 10.5 253.0 44.2 27.67nss-SO2−

4 mg L−1 39 10.5 251.6 43.7 –Lactate mg L−1 13 BDLa 7.8 3.0 1.88Acetate mg L−1 15 BDLa 14.9 4.1 2.57Formate mg L−1 17 0.4 14.4 2.8 1.75Oxalate mg L−1 17 0.6 3.6 1.3 0.81Mn mg L−1 39 0.01 0.28 0.04 0.03Fe mg L−1 39 0.06 3.02 0.40 0.25HCHO mg L−1 39 BDLa 5.9 0.4 0.25H2O2 mg L−1 39 BDLa 3.3 0.8 0.50S(IV) mg L−1 39 BDLa 1.1 0.4 0.25OCc mg L−1 17 BDLa 211.8 37.4 –ECd mg L−1 17 BDLa 8.5 0.3 –Average PM2.5 level µg m−3 39 0.7 81.6 15.9 –

a BDL means below detection limit, b volume weighted mean concentration, c OC means organic carbon, d EC meanselemental carbon.

chromatography (Dionex, IC-2500) (Guo et al., 2012; Yanget al., 2012). Trace metals such as Fe and Mn were analyzedusing inductively coupled plasma mass spectrometry (ICP-MS; Agilent 7500a). The concentrations of organic carbon(OC) and elemental carbon (EC) in cloud water were deter-mined using a thermal–optical transmittance (TOT) carbonanalyzer (Sunset Laboratory, Tigard, OR, USA). For eachcloud sample, a small amount was dropped onto the sur-face of a small standard-size punch (∼ 1.5 cm2) from a pre-combusted quartz filter and analyzed based on the NIOSHprotocol 870 TOT program (Khan et al., 2009; Xu et al.,2017).

2.3 Monitoring of microphysical parameters

A fog monitor (model FM-120, Droplet Measurement Tech-nologies Inc., USA) was used in situ to monitor the LWC, themedian volume diameter (MVD), the effective diameter (ED)and Nd of the cloud droplets with a time resolution of 1 s(for specifications, see Droplet Measurement Technologies,2012). During 24 July to 23 August 2014, 24 cloud eventswere monitored. The measuring range of cloud droplet di-ameters was from 2 to 50 µm, divided into 20 bins. The sam-ple velocity is 15 m s−1 and the sampling flow is 1 m3 min−1.Cloud droplets can not be efficiently collected at low LWC

and Nd values. Based on our experience, the sampling lim-itations associated with LWC and Nd were 0.01 g m−3 and60 cm−3, respectively.

2.4 Measurements of ambient air pollutants andmeteorological parameters

The concentrations of inorganic water-soluble ions, the levelsof PM2.5 and the meteorological parameters were monitoredin real time during the observation periods. The SO2−

4 , NO−3and NH+4 in PM2.5 were measured using two online ion chro-matographs coupled with a wet rotating denuder and a steam-jet aerosol collector (Marga ADI 2080, Applikon-ECN). Abeta attenuation and optical analyzer (model 5030 SHARPmonitor, Thermo Scientific) was used to monitor the levelsof PM2.5. Meteorological parameters including the ambienttemperature, relative humidity, wind speed and wind direc-tion were measured using an automatic meteorological sta-tion.

www.atmos-chem-phys.net/17/9885/2017/ Atmos. Chem. Phys., 17, 9885–9896, 2017


3 Results and discussion

3.1 Chemical properties of cloud water

3.1.1 Acidity

The pH values, electrical conductivity and chemical compo-sitions (inorganic ions, organic acids, metals, HCHO, H2O2,sulfur(IV), OC and EC) of the cloud droplets are summa-rized in Table 1. The pH of the cloud water varied widelyfrom 3.80 to 6.93. The volume-weighted mean (VWM) pHwas 5.87, which is slightly higher than the background pHof 5.6 yielded by CO2 in the atmosphere. The analyzed 39cloud samples were divided into two groups. One contained17 summer samples (i.e., those that were collected from Julyto August) and the other group contained 22 autumn sam-ples (i.e., those that were collected from September to Octo-ber). About 52 % of the summer samples were under pH 5.6and 12 % were under pH 4.5. The corresponding percentagesfor the autumn samples were 14 and 9 %, respectively. Thismeans that some of the cloud samples at Mt. Tai were acidic,especially in the summer. If comparing with other orographicstations less affected by anthropogenic pollution, the VWMpH of clouds at Mt. Tai was higher, as shown in Table 2.Moreover, the VWM pH at Mt. Tai significantly increasedfrom a reported value of 3.86 during 2007–2008 (Guo et al.,2012) to 5.87 in the present study. The detailed reasons forthe big decrease in cloud-water acidity are discussed in Sect.3.1.2.

3.1.2 Chemical composition

The cloud samples contained high concentrations of water-soluble ions at Mt. Tai. The dominant ions were nitrate, sul-fate, ammonium and calcium with VWM concentrations of56.4, 44.2, 34.2 and 5.9 mg L−1, respectively. These ionsrepresented 88.1 % of the total determined ion concentra-tions (TDIC). The concentrations of minor ions includingchloride, potassium, sodium, magnesium and organic acidsranged from 0.7 to 4.1 mg L−1, only amounting to 10.6 % ofthe TDIC. Because of the frequent agricultural and livestockactivities near Mt. Tai, NH+4 was the predominant cation(Cai et al., 2015; Xu et al., 2015). Calcium was the sec-ond most abundant cation and likely originated from sand-storms and/or construction activities. The concentration ofSO2−

4 amounted to 27.7 % of the TDIC, which made SO2−4

the second most abundant anion. The concentration of non-sea-salt sulfate (nss-SO2−

4 ) was calculated using the equation[nss-SO2−

4 ]= [SO2−4 ]− 0.2455[Na+]. In this equation, it is

assumed that the chemical properties of sea salt sulfate (ss-SO2−

4 ) in particles are identical to those in sea water and thesoluble Na+ solely originates from sea salt (Morales et al.,1998). Through calculation, the nss-SO2−

4 represented 93.5–100 % of the total SO2−

4 in this study. Furthermore, this mightbe underestimated because soil dust and biomass combustion

Table2.C

omparison

ofaveragedionic

concentrations(µeq

L−

1)

ofcloudw

atercollectedatM

t.Taiwith

otherregionsin

thew

orld.

SitePeriod

Altitude

pHE

CN

a+

NH+4

K+

Mg 2+

Ca 2+

Cl −

NO−3

SO2−

4R

eference(m

)(µS

cm−

1)

Whiteface

Mountain,

May–Septem

ber20061483

3.8879.6

3.7149.3

2.17.4

26.67.2

79.2220.4

Aleksic

etal.(2009)N

Y,USA

Szrenica,PolandD

ecember2005–

13304.55

80100

21045

49140

93240

200B

łasetal.(2010)

Decem

ber2006M

t.Niesen,

2006–20072362

6.434.4

43143.5

512.6

46.810.6

8772.3

Michna

etal.(2015)Sw

itzerlandSinhagad,India

2007–20101450

686

20428

1717

196234

68198

Budhavantetal.(2014)

Mt.H

eng,China

March–M

ay2009

12793.8

115.2666.03

356.4717.25

5.4929.83

21.07158.8

196.39Sun

etal.(2010)M

t.Tai,China

2007–20081545

3.86–

25.01215

55.133.0

19393.4

4071064

Guo

etal.(2012)M

t.Tai,China

July–October2014

15455.87

169.039.7

1900.832.7

60.5295.5

82.5910.2

920.9(this

work)



are also sources of Na+ besides sea salts (Lu et al., 2010;Sripa et al., 1996). The high ratio of ss-SO2−

4 to SO2−4 in-

dicated that anthropogenic sulfur emissions were the mainsources of SO2−

4 in the cloud samples at Mt. Tai. It shouldbe noted that the VWM of the concentration of SO2−

4 wasalmost the same as that reported during 2007–2008, but theconcentration of NO−3 increased significantly by a factor of2.24 (Guo et al., 2012). This made NO−3 surpass SO2−

4 to bethe predominant anion in 2014. Generally speaking, the pre-vious research indicated that the scavenging of aerosol nitrateand the uptake of gaseous nitric acid are the main sources ofnitrate in cloud/fog water (Collett Jr. et al., 2002). Our re-sults imply that nitrate precursors (mainly NOx from powerplants and/or motor vehicles) had a substantial increase since2007–2008.

Generally, the pH of cloud water is determined by the bal-ance between the acid and the alkaline components. Two fac-tors can decrease the acidity of cloud water: a large input ofalkaline ions and/or a decrease in acid anions. Although theVWM concentration of NO−3 increased significantly, the ad-ditional increases in NH+4 and Ca2+ should also be noted. Es-pecially for NH+4 , the VWM concentration of NH+4 increasedsince 2007–2008 by a factor of 1.56 (Guo et al., 2012). Thismay be attributable to the increasing use of agricultural fer-tilization and soil acidification (Cai et al., 2015; Xu et al.,2015). As a result, the increased levels of NH+4 and Ca2+

play a crucial role in neutralizing the soluble acid ions (NO−3and SO2−

4 ) and decreasing the acidity of cloud water in 2014.The VWM concentrations of acetate, lactate, formate and

oxalate were 4.1, 3.0, 1.75 and 0.81 mg L−1, respectively, ac-counting for 7.01 % of the TDIC. Based on the sources orsource strengths of formic acid and acetic acid, the formic-to-acetic acid ratio (F / A) could be used as an indicator todetermine the sources of organic acids (Sun et al., 2016; Tanet al., 2010). A low ratio indicated the important role of di-rect emissions (such as biomass emission, combustion activ-ities and automobile exhaust) whereas a high ratio indicatedthe in situ photochemical generation of formic acid (Talbotet al., 1988; Tanner and Law, 2003). In the collected cloudsamples, formic acid and acetic acid were highly correlated(r = 0.758, p ≤ 0.01). F / A was about 0.78 (less than 1),suggesting that direct emissions were important sources oforganic acids (Kieber et al., 2002; Li et al., 2011). Oxalicacid was significantly correlated with formic acid (r = 0.667,p ≤ 0.01) and acetic acid (r = 0.638, p ≤ 0.01). This im-plied that formic acid, acetic acid and oxalic acid were proba-bly emitted from the same sources and/or accumulated undersimilar physical conditions (Tanner and Law, 2003). No sig-nificant correlations were found between lactic acid and theother three carboxylic acids; no significant correlations werefound between lactic acid and the other water-soluble ions inthe cloud samples. This implied that the emission source oflactic acid was different from formic, acetic and oxalic acids.

3.2 Microphysical properties of cloud water

3.2.1 Microphysical parameters

The sampling period, number of cloud samples, mean levelof PM2.5, mean microphysical parameters and meteorolog-ical conditions for each cloud event are summarized in Ta-ble 3. There was great diversity in the Nd and the LWCamong the cloud events. The mean values of Nd rangedwidely from 79 to 722 cm−3 and the mean values of LWCranged widely from 0.01 to 0.39 g m−3. Orographic cloud isa highly heterogeneous system consisting of randomly dis-tributed air volumes with different characteristics (Gonser etal., 2012). This feature of orographic cloud generally deter-mines the large differences in CDSD, LWC and aerosol com-position of different cloud events.

3.2.2 Cloud droplet size distribution

The CDSD, which indicates the dynamic and thermody-namic properties of a cloud system, is one of the most cru-cial determinants of the microstructures of cloud (Yin et al.,2011). To investigate the CDSD, four typical cloud events(labeled A, B, C and D) were studied in light of their meanPM2.5 levels of 81.6, 43.0, 25.0 and 11.1 µg m−3 for eventsA, B, C and D, respectively. As shown in Fig. 1, all of thecloud droplets in cloud samples were smaller than 26.0 µm.As the cloud processes continued, droplets ranging from 6.0to 9.0 µm became dominant. The ratios of cloud droplets with6.0–9.0 µm to all droplet sizes were relatively stable amongthe four cloud events (between 0.6 : 1 and 0.7 : 1). The max-imum Nd, which could reach over 1950 cm−3, always oc-curred at a droplet size of 7.0 µm.

An examination of the meteorological parameters withthe microphysical properties of the clouds showed that theLWC somewhat influenced the CDSD. Higher LWC valuesincreased the number of larger cloud droplets and broad-ened the droplet size distribution, while lower LWC valuesinhibited the formation of larger cloud droplets. The forma-tion stage of cloud event B, which occurred during 01:30–02:40 UTC/GMT+8 on 23 August 2014, provided a clear ex-ample. At 02:30, the LWC was relatively low with a valueof 0.09 g m−3. About 8.6 % of the cloud droplets had diame-ters above 10.0 µm and 27.6 % had diameters below 5.0 µm.After 8 min, the LWC sharply increased to 0.29 g m−3. Thecorresponding ratios were 16.3 and 17.1 % for diameters be-low 10.0 and 5.0 µm, respectively. Moreover, cloud dropletslarger than 16.0 µm started to appear and the CDSD changedfrom a monomodal distribution to a weakly bimodal distri-bution. With the development of the cloud event, the stan-dard deviation of the CDSD represented a positive corre-lation with LWC values. It showed that a high LWC couldbroaden the droplet size distribution and increase the rangeof cloud droplets. This situation also occurred in many othercloud events at Mt. Tai.



Table 3. Description of monitored cloud events at Mt. Tai with monitoring times, number of samples (no. samples) and averaged values ofliquid water content (LWC), median volume diameter (MVD), effective diameter (ED), number concentration (Nd), temperature (T ) andrelative humidity (RH).

Event Start Stop No. Duration PM2.5a LWCa Nd

a MVDa EDa T RHa

(UTC/GMT+8) (UTC/GMT+8) samples (h) (µg m−3) (g m−3) (cm−3) (µm) (µm) (◦C) (%)

1 24 July 2014 08:30 24 July 2014 23:20 3 14.8 14.5 0.24 408 12.7 11.0 15.5–22.6 97.92 (D) 25 July 2014 12:00b 25 July 2014 21:40 2 9.7 11.1 0.18 719 8.3 8.3 13.6–14.6 100.03 26 July 2014 23:06 27 July 2014 05:13 0 6.1 100.7 0.04 211 7.8 7.4 15.7–17.0 99.04 (A) 28 July 2014 22:40b 29 July 2014 04:00 1 5.3 81.6 0.09 337 8.2 7.8 16.5–17.6 99.25 29 July 2014 20:33 29 July 2014 22:20 0 1.8 65.6 0.14 694 7.8 7.6 18.5–18.9 99.36 30 July 2014 12:46 30 July 2014 13:50 1 1.1 13.2 0.21 308 12.6 11.8 16.8–18.5 99.57 (C) 30 July 2014 20:20b 30 July 2014 22:40 0 2.3 25.0 0.08 253 9.2 9.2 16.9–18.2 99.68 31 July 2014 19:11 1 August 2014 09:19 2 14.1 20.1 0.18 329 12.6 11.5 17.9–19.1 99.59 4 August 2014 23:42 5 August 2014 11:30 1 11.8 65.8 0.13 539 9.0 8.5 19.5–21.9 99.310 5 August 2014 18:45 6 August 2014 06:13 1 11.5 40.0 0.11 227 11.1 9.8 16.0–20.3 99.311 9 August 2014 07:41 9 August 2014 09:32 0 1.8 17.4 0.06 261 7.9 7.7 13.7–14.0 100.012 11 August 2014 20:42 11 August 2014 21:09 0 0.4 173.3 0.06 392 8.3 7.7 17.6–17.9 99.713 12 August 2014 23:04 13 August 2014 03:55 2 4.8 66.1 0.19 536 9.4 9.1 13.8–16.9 99.014 13 August 2014 18:58 14 August 2014 06:22 3 11.4 34.5 0.19 312 10.9 9.7 13.5–15.9 98.415 14 August 2014 17:35 14 August 2014 19:52 0 2.3 94.6 0.02 104 7.2 6.5 15.7–17.7 98.816 15 August 2014 18:52 16 August 2014 05:59 0 11.1 66.4 0.04 283 6.9 6.5 15.0–17.6 99.217 16 August 2014 19:45 17 August 2014 05:10 0 9.4 93.9 0.03 157 8.3 7.3 15.5–18.2 98.418 17 August 2014 10:02 17 August 2014 11:13 1 1.2 63.5 0.39 722 11.7 10.6 14.9–17.0 99.219 17 August 2014 21:57 18 August 2014 01:23 1 3.4 52.5 0.10 366 8.5 8.3 14.3–15.2 99.120 18 August 2014 08:42 18 August 2014 11:05 0 2.4 – 0.03 118 7.2 6.8 15.0–16.5 98.421 21 August 2014 20:00 22 August 2014 13:48 0 17.8 57.9 0.02 109 7.0 6.5 15.9–20.7 96.322 (B) 23 August 2014 01:30b 23 August 2014 09:20 3 7.8 43.0 0.21 624 9.6 9.4 16.2–17.4 99.623 23 August 2014 18:12 23 August 2014 19:54 0 1.7 70.6 0.01 88 6.8 6.3 16.8–17.8 99.524 25 August 2014 02:25 25 August 2014 06:40 0 4.2 29.4 0.01 79 5.7 5.3 13.8–15.0 97.8

a The arithmetic mean value. b The selected four typical cloud events according to the average PM2.5 level for 28 July 2014 22:40 to 29 July 2014 04:00 (event A, 81.6 µg m−3),23 August 2014 01:30 to 23 August 2014 09:20 (event B, 43.0 µg m−3), 30 July 2014 20:20 to 30 July 2014 22:40 (event C, 25.0 µg m−3) and 25 July 2014 12:00 to 25 July 2014 21:40(event D, 11.1 µg m−3).

3.2.3 Cloud scavenging effect

Cloud processes together with wet deposition play crucialroles in scavenging atmospheric aerosols. Based on the ini-tial PM2.5 levels, cloud processes can be classified into twotypes: type I (including events A and B), which have highinitial PM2.5 levels, and type II (including events C and D),which have low initial PM2.5 levels.

Type I cloud processes have high levels of aerosol scav-enging activity. Using event A as an example, at the begin-ning of the cloud process, there was a relatively high level ofPM2.5 (approximately 128 µg m−3) and Nd increased sharplyfrom 6 to 437 cm−3 over 1 min. As the cloud process contin-ued, the level of PM2.5 decreased and then fluctuated witha mean concentration of 78.2 µg m−3. About 30 min later,the Nd reached a maximum with 1538 cm−3 and the levelof PM2.5 reached a minimum with 23.9 µg m−3, which indi-cated a high PM2.5 removal efficiency of 81.3 %. The some-what inverse relationship between Nd and the level of PM2.5reflects the efficient pollutant removal effect of cloud forma-tion. In type II cloud processes, the levels of PM2.5 were rel-atively low at the initial stage. But for both types of cloudevents, the Nd significantly decreased and the PM2.5 levelsevidently increased as cloud events began to dissipate. Thismay be due to the evaporation of water content that con-

densed on the particles, which freed the CCN and formedhaze. This confirmed that PM2.5 was one of the importanttypes of CCN at Mt. Tai. So, PM2.5 mass concentration wasused as a proxy for CCN number concentration in this study.

3.3 Interaction between aerosols and cloud chemicalproperties

As illustrated in Fig. 2, the TDIC was strongly correlatedwith the levels of PM2.5. High levels of PM2.5 normally leadto high TDIC, whereas low levels of PM2.5 usually lead tolow TDIC. The pH values of cloud samples were somewhataffected by the concentrations of PM2.5. The lower pH val-ues were likely to occur at higher concentrations of PM2.5.Generally, changes to the solute concentrations in cloud wa-ter can be caused by a combination of factors such as themicrophysical conditions, the CCN properties, the chemicalreactions in the cloud droplets and the gas–liquid phase equi-librium (van Pinxteren et al., 2016). Our data emphasize thecrucial effect of PM2.5 on the changes in ion concentrations.PM2.5 is likely to be the main source of ions in cloud water.

To understand the exchange and variation of the threemajor ions (SO2−

4 , NO−3 and NH+4 ) between the aerosolphase and the cloud phase at the summit of Mt. Tai, weanalyzed three cloud samples (CE-Aug23#1 during 02:30–



Figure 1. The variation of wind speed (WS, m s−1), wind direction (WD), temperature (T, ◦C), relative humidity (RH, %), PM2.5 level(µg cm−3), liquid water content (LWC, g m−3), standard deviation of cloud droplet size distribution (SD, µm) and cloud droplet numberconcentration (CDNC, cm−3) during four typical cloud events: (a) event A (28 July 2014 22:40 to 29 July 2014 04:00), (b) event B (23 Au-gust 2014 01:30 to 23 August 2014 09:20), (c) event C (30 July 2014 20:20 to 30 July 2014 22:40) and (d) event D (25 July 2014 12:00 to25 July 2014 21:40).

04:38, CE-Aug23#2 during 04:38–06:21 and CE-Aug23#3during 06:21–09:20) that were collected from the same cloudevent (event B on 23 August 2014). As shown in Fig. 3,in the aerosol phase, the concentrations of SO2−

4 , NO−3 andNH+4 decreased with increases in LWC and vice versa. In thecloud phase, high LWC values meant large cloud dropletsand low concentrations of major ions while low LWC valuesinduced small cloud droplets with high levels of SO2−

4 , NO−3and NH+4 . Elbert et al., (2000) also observed an inverse rela-tionship between the ion concentrations and the LWC values.Between CE-Aug23#1 and CE-Aug23#2, the ion concentra-tions decreased by factors of 2.29, 2.07 and 1.51 for SO2−

4 ,NO−3 and NH+4 , respectively. Meanwhile, the LWC increasedfrom 0.04 to 0.32 g m−3 and the ED increased from 6.7 to10.2 µm. At the dissipation stage of the cloud event, the LWCdecreased to less than 0.10 g m−3 and the ED shrank to about6.6 µm. Simultaneously, the ion concentrations significantly

increased by factors of 1.56, 1.18 and 1.40 for SO2−4 , NO−3

and NH+4 , respectively.The above results demonstrate that cloud water is an im-

portant sink of soluble ions in the atmosphere and smallcloud droplets tend to contain higher concentrations of solu-ble ions than larger ones. SO2−

4 , NO−3 and NH+4 in the aerosolphase were primarily assumed to be transferred to the cloudphase. However, the concentrations of the soluble compo-nents in the cloud phase could not be accurately predictedif only based on their concentrations in the aerosol phase, asthe strong dilution effect of the cloud-water content must alsobe considered. The concentrations of ions in the cloud phasewere primarily determined by two factors: the sources of theions (i.e., the corresponding ion concentrations in the parti-cles acted as CCN) and the LWC values (which representsthe dilution effect of the cloud water). The similar variationtrends of SO2−

4 , NO−3 and NH+4 in both the aerosol phase andthe cloud phase confirmed that LWC was an important factoraffecting the ion concentrations in the cloud water at Mt. Tai



Figure 2. Ion and organic acid concentrations (µeq L−1) with the variation of PM2.5 levels (µg cm−3) and pH of cloud-water samples.

Figure 3. Variation trend of hour-averaged LWC (g m−3), ED (µm) and the concentrations of NO−3 , SO2−4 and NH+4 in the aerosol phase

and the cloud phase during the cloud event on 23 August 2014.



(Aleksic and Dukett, 2010; Elbert et al., 2000). As mentionedabove, LWC also determined the size of cloud droplets. Thisultimately showed that high concentrations of soluble ionsare concentrated in small cloud droplets. It should be notedthat, compared with SO2−

4 and NO−3 , the concentration ofNH+4 in the aerosol phase did not directly increase at the dis-sipation stage of the cloud event. This was primarily due tothe high solubility of NH3, which dissolved in the cloud wa-ter and gave rise to the increase in the concentration of NH+4in cloud samples.

3.4 Water-soluble ions and droplet size under PM2.5

Secondary inorganic compounds especially ammonium sul-fate and ammonium nitrate were the main hygroscopic com-pounds of particulate matter. The presence of these com-pounds could enhance the hygroscopic ability of atmosphericparticles and facilitate their ability to act as CCN (Wang etal., 2014; Ye et al., 2011). These water-soluble ions are pri-marily transferred to the cloud phase during the formationof cloud droplets by activation of CCN. As mentioned be-fore, SO2−

4 , NO−3 , NH+4 and Ca2+ were the most predomi-nant ions in cloud samples collected at Mt. Tai. The averagedconcentrations surpassed 88.1 % of the TDIC. Presumably,PM2.5 was the main source of the mentioned soluble ions incloud water. In order to investigate the trend between water-soluble ions and cloud droplet size under different PM2.5 lev-els, 17 cloud samples collected from 25 July to 23 Augustwere studied as shown in Fig. 4. As can be seen, high PM2.5levels represented high ion concentrations and small clouddroplets. This confirmed again that PM2.5 acting as CCN wasthe main source of soluble ions in cloud water. High PM2.5levels would lead to a large source of CCN, increase the com-petition of ambient water vapor and hinder the formation oflarge cloud droplets.

It should be noticed that sometimes the Nd varied at thesame PM2.5 level in Fig. 4b. This was caused by the varia-tion of LWC values or ambient RH in different monitoringmoments (Ackerman et al., 2004). The low RH (representinginsufficient water vapor in the atmosphere) would impede thehygroscopic growth of particles and the activation of droplets(Gonser et al., 2012; Liu et al., 2011), affecting the formationof cloud droplets.

4 Conclusions

In 2014, samples of clouds at Mt. Tai showed that the VWMpH of the cloud samples was 5.87. It is much higher thanthat reported by previous studies taking place at the same siteduring 2007–2008. The cloud water contained much higherconcentrations of ions than the samples collected at otherorographic sites, indicating the strong influence of anthro-pogenic emissions on clouds at the summit of Mt. Tai. Thedominant ion species were NH+4 , SO2−

4 , Ca2+ and NO−3 ,

Figure 4. (a) The variation of the ED of cloud droplets and the sumof four water-soluble ions (SO2−

4 , NO−3 , NH+4 and Ca2+) underdifferent PM2.5 levels. (b) The variation of the ED and Nd of clouddroplets under different PM2.5 levels.

which amounted to more than 88.1 % of the TDIC. TheNO−3 content of the cloud water was significantly higherthan that during 2007–2008. However, the increase in theNH+4 concentration (mainly from NH3) exceeded that ofNO−3 (mainly from NOx), leading to net neutralization andreduced cloud acidity. The rapid increase in the concentra-tion of NH+4 should be attributable to agricultural fertiliza-tion and soil acidification which frequently occurred duringrecent years (Cai et al., 2015; Xu et al., 2016). The micro-physical parameters of the cloud samples varied enormouslybetween the cloud events. The cloud droplets were all smallerthan 26.0 µm and most were 6.0–9.0 µm. The maximum Ndwas associated with droplet sizes of 7.0 µm. Higher LWCvalues could facilitate the formation of larger cloud dropletsand broadened the droplet size distribution. A strong interac-



Figure 5. A schematic of aerosol particles’ impact on the cloud droplet sizes. (a) PM2.5 > 35 µg m−3, (b) PM2.5 ≤ 35 µg m−3.

tion was observed between the concentrations of soluble ionsin cloud droplets and the levels of PM2.5 in the atmosphere.The clouds played a crucial role in scavenging atmosphericaerosols. Higher PM2.5 levels resulted in higher TDIC. Thelower pH values were likely to occur at higher PM2.5 concen-trations. We found that the dilution effect of cloud water wasstrong and it should not be ignored when estimating concen-trations of soluble components in the cloud phase.

In summary, the mechanism of cloud droplet formation issummarized in Fig. 5. According to the concentrations ofPM2.5, cloud events were divided into two categories: onewas the PM2.5 concentrations greater than 35 µg m−3 andthe other was the PM2.5 concentrations less than or equal to35 µg m−3. Cloud droplets would be formed on condensationnuclei (usually aerosols including secondary aerosol, dust,sea salt, etc.) through water vapor condensation and thenundergo hygroscopic growth. The soluble ions in conden-sation nuclei and ambient gases could enter cloud dropletsthrough surface reactions and consequently participate indissolution, diffusion, dilution and aqueous reactions in thecloud phase. Higher aerosol concentrations supplied higherconcentrations of soluble ions for cloud droplets and facili-tated the formation of smaller sizes of cloud droplets, which

caused the high concentrations of soluble ions in small clouddroplets.

Data availability. The data for studying chemical and microphysi-cal properties of cloud in this article are available from the authorsupon request ([email protected]).

Competing interests. The authors declare that they have no conflictof interest.

Special issue statement. This article is part of the special issue“Regional transport and transformation of air pollution in easternChina”. It is not associated with a conference.

Acknowledgements. This work was supported by the TaishanScholar grant (ts20120552), the National Natural Science Foun-dation of China (41375126, 41275123, 21190053, 21177025),the Cyrus Tang Foundation (no. CTF-FD2014001), the Min-istry of Science and Technology of China (2016YFC0202701,2014BAC22B01), the Strategic Priority Research Program of



the Chinese Academy of Sciences (grant no. XDB05010200),and the Natural Science Foundation of Shandong Province(no. ZR2014BQ031).

Edited by: Hang SuReviewed by: two anonymous referees

References

Ackerman, A. S., Kirkpatrick, M. P., Stevens, D. E., and Toon, O.B.: The impact of humidity above stratiform clouds on indirectaerosol climate forcing, Nature, 432, 1014–1017, 2004.

Aleksic, N. and Dukett, J. E.: Probabilistic relationship between liq-uid water content and ion concentrations in cloud water, Atmos.Res., 98, 400–405, 2010.

Aleksic, N., Roy, K., Sistla, G., Dukett, J., Houck, N., and Casson,P.: Analysis of cloud and precipitation chemistry at WhitefaceMountain, NY, Atmos. Environ., 43, 2709–2716, 2009.

Błas, M., Sobik, M., and Twarowski, R.: Changes of cloud waterchemical composition in the Western Sudety Mountains, Poland,Atmos. Res., 87, 224–231, 2008.

Błas, M., Polkowska, Z., Sobik, M., Klimaszewska, K., Nowinski,K., and Namiesnik, J.: Fog water chemical composition in dif-ferent geographic regions of Poland, Atmos. Res., 95, 455–469,2010.

Borys, R. D., Lowenthal, D. H., and Mitchell, D. L.: The relation-ships among cloud microphysics, chemistry, and precipitationrate in cold mountain clouds, Atmos. Environ., 34, 2593–2602,2000.

Budhavant, K. B., Rao, P. S. P., Safai, P. D., Granat, L., and Rodhe,H.: Chemical composition of the inorganic fraction of cloud-water at a high altitude station in West India, Atmos. Environ.,88, 59–65, 2014.

Cai, Z., Wang, B., Xu, M., Zhang, H., He, X., Zhang, L., and Gao,S.: Intensified soil acidification from chemical N fertilization andprevention by manure in an 18-year field experiment in the redsoil of southern China, J. Soils Sediments, 15, 260–270, 2015.

Collett Jr., J. L., Hoag, K. J., Sherman, D. E., Bator, A., andRichards, L. W.: Spatial and temporal variations in San JoaquinValley fog chemistry, Atmos. Environ., 33, 129–140, 1998.

Collett Jr., J. L., Bator, A., Sherman, D. E., Moore, K. F., Hoag,K. J., Demoz, B. B., Rao, X., and Reilly, J. E.: The chemicalcomposition of fogs and intercepted clouds in the United States,Atmos. Res., 64, 29–40, 2002.

Demoz, B. B., Collett Jr., J. L., and Daube Jr., B. C.: On the CaltechActive Strand Cloudwater Collectors, Atmos. Res., 41, 47–62,1996.

DROPLET MEASUREMENT TECHNOLOGIES, I.: ParticleAnalysis and Display System (PADS) 3.6.3 Overview ManualDOC-0300, Rev C-11, 2012.

Elbert, W., Hoffmann, M. R., Krämer, M., Schmitt, G., and An-dreae, M. O.: Control of solute concentrations in cloud and fogwater by liquid water content, Atmos. Environ., 34, 1109–1122,2000.

Gonser, S. G., Klemm, O., Griessbaum, F., Chang, S. C., Chu, H.S., and Hsia, Y. J.: The Relation Between Humidity and LiquidWater Content in Fog: An Experimental Approach, Pure Appl.Geophys., 169, 1–13, 2012.

Gultepe, I. and Milbrandt, J. A.: Microphysical Observations andMesoscale Model Simulation of a Warm Fog Case during FRAMProject, Pure Appl. Geophys., 164, 1161–1178, 2007.

Guo, J., Wang, Y., Shen, X., Wang, Z., Lee, T., Wang, X., Li, P., Sun,M., Collett Jr., J. L., Wang, W., and Wang, T.: Characterization ofcloud water chemistry at Mount Tai, China: Seasonal variation,anthropogenic impact, and cloud processing, Atmos. Environ.,60, 467–476, 2012.

Herckes, P., Lee, T., Trenary, L., Kang, G., Chang, H., and CollettJr., J. L.: Organic matter in central California radiation fogs, En-viron. Sci. Technol., 36, 4777–4782, 2002.

Khan, A. J., Swami, K., Ahmed, T., Bari, A., Shareef, A., and Hu-sain, L.: Determination of elemental carbon in lake sediments us-ing a thermal–optical transmittance (TOT) method, Atmos. Env-iron., 43, 5989–5995, 2009.

Kieber, R. J., Peake, B., Willey, J. D., and Avery, G. B.: Dissolvedorganic carbon and organic acids in coastal New Zealand rain-water, Atmos. Environ., 36, 3557–3563, 2002.

Kim, M.-G., Lee, B.-K., and Kim, H.-J.: Cloud/Fog Water Chem-istry at a High Elevation Site in South Korea, J. Atmos. Chem.,55, 13–29, 2006.

Lee, A. K. Y., Hayden, K. L., Herckes, P., Leaitch, W. R., Lig-gio, J., Macdonald, A. M., and Abbatt, J. P. D.: Characteri-zation of aerosol and cloud water at a mountain site duringWACS 2010: secondary organic aerosol formation through ox-idative cloud processing, Atmos. Chem. Phys., 12, 7103–7116,https://doi.org/10.5194/acp-12-7103-2012, 2012.

Li, P., Li, X., Yang, C., Wang, X., Chen, J., and Collett Jr., J. L.: Fogwater chemistry in Shanghai, Atmos. Environ., 45, 4034–4041,2011.

Liu, P. F., Zhao, C. S., Göbel, T., Hallbauer, E., Nowak, A., Ran,L., Xu, W. Y., Deng, Z. Z., Ma, N., Mildenberger, K., Henning,S., Stratmann, F., and Wiedensohler, A.: Hygroscopic propertiesof aerosol particles at high relative humidity and their diurnalvariations in the North China Plain, Atmos. Chem. Phys., 11,3479–3494, https://doi.org/10.5194/acp-11-3479-2011, 2011.

Lu, C., Niu, S., Tang, L., Lv, J., Zhao, L., and Zhu, B.: Chemicalcomposition of fog water in Nanjing area of China and its relatedfog microphysics, Atmos. Res., 97, 47–69, 2010.

Marinoni, A., Laj, P., Sellegri, K., and Mailhot, G.: Cloudchemistry at the Puy de Dôme: variability and relationshipswith environmental factors, Atmos. Chem. Phys., 4, 715–728,https://doi.org/10.5194/acp-4-715-2004, 2004.

Michna, P., Werner, R. A., and Eugster, W.: Does fog chemistryin Switzerland change with altitude?, Atmos. Res., 151, 31–44,2015.

Miles, N. L., Verlinde, J., and Clothiaux, E. E.: Cloud Droplet SizeDistributions in Low-Level Stratiform Clouds, J. Atmos. Sci., 57,295–311, 2000.

Moore, K. F., Sherman, D. E., Reilly, J. E., Hannigan, M. P., Lee,T., and Collett, J. L.: Drop size-dependent chemical composi-tion of clouds and fogs. Part II: Relevance to interpreting theaerosol/trace gas/fog system, Atmos. Environ., 38, 1403–1415,2004.

Morales, J. A., Pirela, D., Nava, M. G. D., Borrego, B. Z. S. D.,Velásquez, H., and Durán, J.: Inorganic water soluble ions in at-mospheric particles over Maracaibo Lake Basin in the westernregion of Venezuela, Atmos. Res., 46, 307–320, 1998.


https://doi.org/10.5194/acp-12-7103-2012

https://doi.org/10.5194/acp-11-3479-2011

https://doi.org/10.5194/acp-4-715-2004


Ogawa, N., Kikuchi, R., Okamura, T., Inotsume, J., Adzuhata, T.,Ozeki, T., and Kajikawa, M.: Evaluation of ionic pollutants incloud droplets at a mountain ridge in northern Japan using con-strained oblique rotational factor analysis, Atmos. Res., 54, 279–283, 2000.

Portin, H., Leskinen, A., Hao, L., Kortelainen, A., Miettinen, P.,Jaatinen, A., Laaksonen, A., Lehtinen, K. E. J., Romakkaniemi,S., and Komppula, M.: The effect of local sources on particle sizeand chemical composition and their role in aerosol–cloud inter-actions at Puijo measurement station, Atmos. Chem. Phys., 14,6021–6034, https://doi.org/10.5194/acp-14-6021-2014, 2014.

Ravishankara, A. R.: Heterogeneous and Multiphase Chemistry inthe Troposphere, Science, 276, 1058–1065, 1997.

Schell, D., Wobrock, W., Maser, R., Preiss, M., Jaeschke, W.,Georgii, H. W., Gallagher, M. W., Bower, K. N., Beswick, K.M., and Pahl, S.: The size-dependent chemical composition ofcloud droplets, Atmos. Environ., 31, 2561–2576, 1997.

Sripa, P., Tongraar, A., and Kerdcharoen, T.: Chemical compositionof fogwater in an urban area: Strasbourg (France), Environ. Pol-lut., 94, 345–354, 1996.

Sun, M., Wang, Y., Wang, T., Fan, S., Wang, W., Li, P.,Guo, J., and Li, Y.: Cloud and the corresponding precipita-tion chemistry in south China: Water-soluble components andpollution transport, J. Geophys. Res.-Atmos., 115, D22303,https://doi.org/10.1029/2010JD014315, 2010.

Sun, X., Wang, Y., Li, H., Yang, X., Sun, L., Wang, X., Wang, T.,and Wang, W.: Organic acids in cloud water and rainwater at amountain site in acid rain areas of South China, Environ. Sci.Pollut. R., 23, 9529, https://doi.org/10.1007/s11356-016-6038-1,2016.

Talbot, R. W., Beecher, K. M., Harriss, R. C., and Iii, W. R. C.: At-mospheric Geochemistry of Formic and Acetic Acids at a Mid-Latitude Temperate Site, J. Geophys. Res.-Atmos., 93, 1638–1652, 1988.

Tan, Y., Carlton, A. G., Seitzinger, S. P., and Turpin, B. J.: SOAfrom methylglyoxal in clouds and wet aerosols: Measurementand prediction of key products, Atmos. Environ., 44, 5218–5226,2010.

Tanner, P. A. and Law, P. T.: Organic Acids in the Atmosphere andBulk Deposition of Hong Kong, Water Air Soil Pollut., 142, 279–297, 2003.

van Pinxteren, D., Fomba, K. W., Mertes, S., Müller, K., Spindler,G., Schneider, J., Lee, T., Collett, J. L., and Herrmann, H.: Cloudwater composition during HCCT-2010: Scavenging efficiencies,solute concentrations, and droplet size dependence of inorganicions and dissolved organic carbon, Atmos. Chem. Phys., 16,3185–3205, https://doi.org/10.5194/acp-16-3185-2016, 2016.

Wang, X., Ye, X., Chen, H., Chen, J., Yang, X., and Gross, D.S.: Online hygroscopicity and chemical measurement of urbanaerosol in Shanghai, China, Atmos. Environ., 95, 318–326, 2014.

Wang, Y., Wai, K. M., Gao, J., Liu, X., Wang, T., and Wang, W.: Theimpacts of anthropogenic emissions on the precipitation chem-istry at an elevated site in North-eastern China, Atmos. Environ.,42, 2959–2970, 2008.

Wang, Y., Guo, J., Wang, T., Ding, A., Gao, J., Yang, Z., Collett, J.L., and Wang, W.: Influence of regional pollution and sandstormson the chemical composition of cloud/fog at the summit of Mt.Taishan in northern China, Atmos. Res., 99, 434–442, 2011.

Watanabe, K., Honoki, H., Iwai, A., Tomatsu, A., Noritake,K., Miyashita, N., Yamada, K., Yamada, H., Kawamura, H.,and Aoki, K.: Chemical Characteristics of Fog Water at Mt.Tateyama, Near the Coast of the Japan Sea in Central Japan, Wa-ter, Air, Soil Pollut. 211, 379–393, 2010.

Xu, P., Liao, Y. J., Lin, Y. H., Zhao, C. X., Yan, C. H., Cao,M. N., Wang, G. S., and Luan, S. J.: High-resolution in-ventory of ammonia emissions from agricultural fertilizer inChina from 1978 to 2008, Atmos. Chem. Phys., 16, 1207–1218,https://doi.org/10.5194/acp-16-1207-2016, 2016.

Xu, W., Wang, F., Li, J., Tian, L., Jiang, X., Yang, J., and Chen,B.: Historical variation in black carbon deposition and sources toNorthern China sediments, Chemosphere, 172, 242–248, 2017.

Yang, L., Zhou, X., Wang, Z., Zhou, Y., Cheng, S., Xu, P., Gao,X., Nie, W., Wang, X., and Wang, W.: Airborne fine particu-late pollution in Jinan, China: Concentrations, chemical compo-sitions and influence on visibility impairment, Atmos. Environ.,55, 506–514, 2012.

Ye, X., Ma, Z., Hu, D., Yang, X., and Chen, J.: Size-resolved hygro-scopicity of submicrometer urban aerosols in Shanghai duringwintertime, Atmos. Res., 99, 353–364, 2011.

Yin, J., Wang, D., and Zhai, G.: Long-term in situ measurements ofthe cloud-precipitation microphysical properties over East Asia,Atmos. Res., 102, 206–217, 2011.

Zapletal, M., Kunák, D., and Chroust, P.: Chemical characterizationof rain and fog water in the Cervenohorske Sedlo (Hruby JesenikMountains, Czech Republic), Water Air Soil Pollut., 186, 85–96,2007.

Zimmermann, F., Lux, H., Maenhaut, W., Matschullat, J., Plessow,K., Reuter, F., and Wienhaus, O.: A review of air pollution andatmospheric deposition dynamics in southern Saxony, Germany,Central Europe, Atmos. Environ., 37, 671–691, 2003.

Zipori, A., Rosenfeld, D., Tirosh, O., Teutsch, N., and Erel, Y.: Ef-fects of aerosol sources and chemical compositions on cloud dropsizes and glaciation temperatures, J. Geophys. Res.-Atmos., 120,9653–9669, 2015.


https://doi.org/10.5194/acp-14-6021-2014

https://doi.org/10.1029/2010JD014315

https://doi.org/10.1007/s11356-016-6038-1

https://doi.org/10.5194/acp-16-3185-2016

https://doi.org/10.5194/acp-16-1207-2016

chemical composition and droplet size distribution of ... · 1environment research institute,...

Documents